Project Summary Genome-wide patterns of sequence divergence over evolutionary time provide a unique window into the fundamental metabolic costs imposed on cellular life. Purifying selection eliminates mutations that increase costs and/or promote genetic disease. This process can discern minor cost differences, even ones that may not be readily measurable by direct laboratory experiments. This project will identify and interpret the evolutionary signals imprinted into genomes by one specific cost, the cost of erroneously translating proteins. Mistranslation events lead to protein misfolding, misfolded proteins can be cytotoxic or require costly cleanup, and selection operates both on the codon and on the amino-acid level to minimize cellular exposure to misfolded proteins. The working hypothesis for this project is that among the costs linked to translation, mistranslation-induced misfolding is the dominant one, whereas other costs, including mistranslation-induced loss of function and translation at reduced speed, play a minor role. This hypothesis will be tested using a combination of bioinformatics, mathematical modeling, and computer simulation, and the relative importance of the various translation- linked costs will be quantified. There are three specific aims. 1. What makes translationally optimal codons optimal? 2. Does selection against protein misfolding shapes synonymous codon usage? 3. How does protein biophysics interact with translational selection to constrain sequence evolution?